CN111457927B - Unmanned cruise ship multi-target path planning method under dynamic barrier - Google Patents
Unmanned cruise ship multi-target path planning method under dynamic barrier Download PDFInfo
- Publication number
- CN111457927B CN111457927B CN202010340555.6A CN202010340555A CN111457927B CN 111457927 B CN111457927 B CN 111457927B CN 202010340555 A CN202010340555 A CN 202010340555A CN 111457927 B CN111457927 B CN 111457927B
- Authority
- CN
- China
- Prior art keywords
- node
- wolf
- path
- point
- grid
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
- G01C21/203—Specially adapted for sailing ships
Abstract
The invention discloses a multi-target path planning method for an unmanned cruise ship under a dynamic barrier, and relates to the technical field of water quality sampling and path planning. Firstly, an unmanned aerial vehicle collects an image of a lake surface environment, grid segmentation is carried out, and a starting point and a plurality of sampling points are arranged on a grid map; sequentially optimizing a plurality of sampling points by adopting an improved wolf optimization algorithm, and marking the sampling points in the optimal sequence on a map one by one; calculating an optimal raster path between every two sampling points marked in the raster map by using a D-Lite algorithm to obtain an optimal path from a starting point to a final sampling point; and finally, the autonomous cruise ship cruises along the optimal path. The method improves the convergence factor in the gray wolf optimization algorithm, balances the global search and local search capabilities of the gray wolf optimization algorithm, improves the convergence speed and stability of the gray wolf optimization algorithm, and can realize path planning of a plurality of target points under the condition of a dynamic unknown environment.
Description
Technical Field
The invention belongs to the technical field of lake basin water quality sampling and path planning, and particularly relates to a multi-target path planning method for an unmanned cruise ship under a dynamic barrier.
Background
Nowadays, the problem of water environment protection is more and more emphasized by the nation. Along with the development of science and technology, in the water sampling field, intelligence water sampling cruise ship replaces artifical the collection gradually, uses more and more extensively.
However, the problem of path planning of an intelligent water quality sampling cruise ship has been a key point and a difficulty which are concerned by scholars at home and abroad, and whether the selection of the path reasonably and directly influences the operation efficiency of the cruise ship. The method is suitable for the current environment and can rapidly plan the optimal sampling path, so that the sampling efficiency can be remarkably improved, and the sampling cost can be reduced.
Currently, many traditional methods and intelligent methods are used to solve the problem of multi-target point path planning of cruise ships, such as: genetic algorithms and ant colony algorithms, etc.
The gray wolf optimization algorithm is a group intelligent optimization algorithm proposed in 2014, and can realize sequential optimization of a plurality of target points. However, the algorithm cannot solve the path planning problem of the dynamic unknown environment, and the convergence of the algorithm is slow.
The D × Lite algorithm is an algorithm capable of solving the path planning problem of the dynamic unknown environment, and adopts a reverse search mode. However, the algorithm can only solve the problem of a single target point, and cannot only be used to traverse multiple target points.
Disclosure of Invention
The invention provides a multi-target path planning method for an unmanned cruise ship under a dynamic barrier aiming at the problem of the requirement of collecting multiple points of an autonomous cruise ship in a dynamic environment, and the method utilizes a D × Lite algorithm in combination with an improved Hui wolf optimization algorithm, fully utilizes the advantages of the two algorithms, shortens the time consumed by searching the path, reduces the equipment cost, and effectively solves the path planning problem of multiple target points under the dynamic environment.
The unmanned cruise ship multi-target path planning method under the dynamic barrier comprises the following specific steps:
the method comprises the following steps: the unmanned aerial vehicle collects images of the lake surface environment, and the images are segmented by adopting a grid method to obtain a grid map and set a starting point and d sampling points;
the method specifically comprises the following steps: firstly, dividing a lake surface image into grids with equal size to form a grid map; setting the obstacles in the environment as black grids, filling irregular obstacles, setting the feasible region as a white grid, and setting a starting point and d sampling points in a grid map by using two-dimensional coordinates.
d is a positive integer.
Step two: and (4) carrying out sequential optimization on the d sampling points by adopting an improved wolf optimization algorithm to obtain an optimal sequence and storing the optimal sequence.
The steps of obtaining the optimal sequence by adopting the improved wolf optimization algorithm are as follows:
step 201: parameters of the gray wolf optimization algorithm are initialized.
The parameters include: the number n of the grey wolfs population, the maximum iteration number t, the number of sampling points defined as the dimension d of a search space, the initial position of the grey wolf population, a random path of each grey wolf, and the like;
step 202: searching a prey in the space of d sampling points by each wolf, and constructing a space domain matrix P of the position relation between the search space and the wolf;
first, define the location X of the ith wolfiIs a group of positive integer sequences different from each other, and the formula is as follows:
then
Represents the value of the ith grey wolf in the grey wolf population on the d sampling point;
then, the positions of n wolfs constitute a spatial domain matrix P, the formula is as follows:
in the formula (I), the compound is shown in the specification,
indicating the location of the ith wolf at the mth sample point, each row in the spatial domain matrix P represents the path of a wolf.
Step 203: constructing a distance matrix W to represent the distance between each sampling point and other sampling points;
the formula is as follows:
s(n,m)is the distance between the nth and mth sample points.
Step 204: constructing a fitness function f by using the space domain matrix P and the distance matrix W;
the expression is as follows:
Pirefers to the ith row of wolf-only path in the spatial domain matrix P;
means that in the ith wolf path, the distances between every two sampling points are added and summed.
Step 205: solving a fitness function by using a gray wolf optimization algorithm to obtain an optimal traversal sequence;
the process is as follows:
firstly, the searching process:
in the t-generation search process, the position of the gray wolf is X (t), and the position of the prey is XP(t), the distance D of the wolf from the prey is expressed as:
t denotes the current number of iterations, r2Is a random number from 0 to 1; c is a radical of r2The coefficient of variation.
Then the bounding process:
in the process that the gray wolf colony surrounds the prey, a relationship model of the gray wolf and the prey is constructed, and the formula is as follows:
wherein A and D are the surrounding step length, a is the improved convergence factor, r1Is a random number from 0 to 1, tmaxDenotes the maximum number of iterations, μ1And mu2Is a constant, λ1And λ2For adjusting the coefficients, B is a non-linear continuous function Beta.
Next, the location update procedure:
according to the relation model of gray wolf and prey, the final positions X of three wolfs of alpha, beta and delta are respectively obtained1，X2And X3Under the guidance of alpha, beta and delta, the whole wolf group gradually approaches to the prey, and the target position is continuously determined through the position updating of alpha, beta and delta wolfs, so that the final attack position of the gray wolf on the prey is realized.
In the formula, Xα，Xβ，XδRepresents the updated positions of three wolfs of alpha, beta and delta, A1，A2，A3Three random numbers are shown, and X represents the ultimate attack site of the wolf on the game. A. the1·Dα，A2·Dβ，A3·DδRepresenting the surrounding step sizes of three wolfs, alpha, beta and delta.
Finally, judging whether the maximum iteration times is reached, if so, outputting the optimal sequence in the iteration process; otherwise, sequentially storing three optimal solutions obtained when the fitness function is minimum, and updating the positions of three wolfs, namely alpha, beta and delta.
Step three: according to the optimal sequence obtained by the gray wolf algorithm, marking each sampling point on a grid map one by one;
step four: calculating an optimal raster path between every two sampling points marked in the raster map by using a D-Lite algorithm to obtain an optimal path from a starting point to a final sampling point;
step five: and the autonomous cruise ship starts cruising from the starting point corresponding to the grid map, sequentially passes through each sampling point along the optimal grid path and finally returns to the starting point to finish cruising.
The invention has the advantages that:
(1) a multi-target path planning method for an unmanned cruise ship under a dynamic barrier converts lake environment into a two-dimensional coordinate grid map, simplifies the map and shortens the path planning time of the intelligent water quality sampling cruise ship.
(2) A multi-target path planning method for an unmanned cruise ship under a dynamic barrier is based on an improved Hui wolf optimization algorithm, solves the problem of sequential optimization of a plurality of target points, is more suitable for actual problem operation, enhances the convergence capability of the algorithm by improving the convergence factor, and directly improves the efficiency of water sampling of the cruise ship.
(3) A multi-target path planning method for an unmanned cruise ship under a dynamic obstacle adopts a D-Lite algorithm, meets the path planning requirement of the cruise ship in a dynamic unknown environment, and effectively solves the path planning problem of a plurality of target points by combining with an improved Hui wolf optimization algorithm.
Drawings
FIG. 1 is a flow chart of a multi-objective path planning method for an unmanned cruise ship under a dynamic obstacle according to the present invention;
FIG. 2 is a flow chart of an improved gray wolf optimization algorithm of the present invention;
fig. 3 is a flow chart of D × Lite algorithm path planning in the present invention.
Detailed Description
The present invention will be described in detail with reference to the accompanying drawings.
The invention provides a multi-target path planning method for an unmanned cruise ship under a dynamic barrier, which adopts a D-Lite algorithm and an improved wolf optimization algorithm to collect a plurality of target points for a water quality sampling cruise ship to plan an optimal path, and specifically comprises the following steps as shown in figure 1:
the method comprises the following steps: the unmanned aerial vehicle collects images of the lake surface environment, and the images are segmented by adopting a grid method to obtain a grid map and set a starting point and d sampling points;
the method specifically comprises the following steps: firstly, dividing a lake surface image into grids with equal size to form a grid map; setting the obstacles in the environment as black grids, filling irregular obstacles, setting the feasible region as a white grid, and selecting and setting a starting point and d sampling points in a grid map by using two-dimensional coordinates.
d is a positive integer.
Step two: and D sampling points are sequentially optimized by adopting an improved wolf optimization algorithm, and the optimal sequence of the starting point passing through the sampling points and returning to the starting point is obtained and stored.
As shown in fig. 2, the steps of obtaining the optimal sequence by using the improved grayish optimization algorithm are as follows:
step 201: parameters of the gray wolf optimization algorithm are initialized.
The parameters include: the number n of the wolf populations is 15, the maximum iteration time t is 500, the number of sampling points is defined as 10 for the dimension d of a search space, the initial position of the wolf populations, a random path of each wolf and the like;
step 202: searching a prey in the space of d sampling points by each wolf, and constructing a space domain matrix P of the position relation between the search space and the wolf;
first, define the location X of the ith wolfiIs a group of positive integer sequences different from each other, and the formula is as follows:
then
Represents the value of the ith grey wolf in the grey wolf population on the d sampling point;
then, n wolfs search for the prey in d-dimensional space, the position constitutes the space domain entity matrix P, and the formula is as follows:
in the formula (I), the compound is shown in the specification,
the location of the ith grey wolf at the mth sampling point is shown, the first row in the spatial domain matrix P shows the location sequence of the 1 st grey wolf in the search space, the last column of the matrix shows n grey wolfs in the d-dimensional search space, and each row represents the path of one wolf.
Step 203: constructing a distance matrix W to represent the distance between each sampling point and other sampling points;
the formula is as follows:
s(n,m)is the distance between the nth and mth sample points.
Step 204: constructing a fitness function f by using the space domain matrix P and the distance matrix W; the essence of solving the shortest path problem is to find an optimal grey wolf population position, so that the fitness function is minimum,
the expression is as follows:
Pirefers to the ith row of wolf-only path in the spatial domain matrix P;
means that in the ith wolf path, the distances between every two sampling points are added and summed. For example in the gray wolf position PiIn the order of (2,3,1,5,4), the distance matrix corresponds to s(2,3)、s(3,1)、s(1,5)、s(5,4)、s(4,2)The corresponding fitness function f is sum(s)(2,3),s(3,1),s(1,5),s(5,4),s(4,2))。
Step 205: solving a fitness function by using a gray wolf optimization algorithm to obtain an optimal traversal sequence;
the process is as follows:
firstly, the searching process:
an important criterion of the gray wolf colony in searching the prey is the distance between the gray wolf colony and the prey, wherein in the t generation searching process, the position of the gray wolf is X (t), and the position of the prey is XP(t), the distance D of the wolf from the prey is expressed as:
t denotes the current number of iterations, r2Is a random number from 0 to 1; c is a radical of r2The coefficient of variation.
Then the bounding process:
in the process that the gray wolf colony surrounds the prey, a relationship model of the gray wolf and the prey is constructed, and the formula is as follows:
wherein A and D are the surrounding step length, a is the improved convergence factor, r1Is a random number from 0 to 1, tmaxDenotes the maximum number of iterations, μ1Is constant 0.01, mu2Is constant 0.1, lambda1To adjust the coefficient 1, lambda2For adjustment of the coefficient 0.1, B is a non-linear continuous function Beta function.
The improved convergence factor a has a good curve trend, the convergence factor a can be reduced nonlinearly along with the increase of the iteration times, when the iteration is just started, the convergence factor a can be rapidly reduced nonlinearly, the global search capability in the early stage of the gray wolf optimization algorithm is improved, and along with the increase of the iteration times, the reduction curve of the convergence factor a is similar to linearity, the local search capability in the later stage of the gray wolf optimization algorithm is ensured, and the convergence capability and the stability of the gray wolf optimization algorithm are improved on the whole.
Next, the location update procedure:
according to the relation model of gray wolf and prey, the final positions X of three wolfs of alpha, beta and delta are respectively obtained1，X2And X3Under the guidance of alpha, beta and delta, the whole wolf group gradually approaches to the prey, and the target position is continuously determined through the position updating of alpha, beta and delta wolfs, so that the final attack position of the gray wolf on the prey is realized.
In the formula, Xα，Xβ，XδRepresents the updated positions of three wolfs of alpha, beta and delta, A1，A2，A3Three random numbers are shown, and X represents the ultimate attack site of the wolf on the game. A. the1·Dα，A2·Dβ，A3·DδRepresenting the surrounding step sizes of three wolfs, alpha, beta and delta.
The positions of the wolfs are expressed by the sequence of the sampling points, so the wolfs are integers, but decimals are inevitably generated in the iterative calculation process of the wolfs, and the situation that the iteration cannot be continued occurs. Therefore, after each iteration, the position of the wolf needs to be approximated, the formula is as follows:
[C,I]＝sort(P,2)
where I is an array of size (P) and is the row or column position of the returned sorted element in the original array.
Finally, whether the maximum iteration number reaches the upper limit of 1000 is judged, if so, the optimal sequence in the iteration process and the optimal fitness function f in the iteration process are outputmin(ii) a Otherwise, sequentially storing three optimal solutions obtained when the fitness function is minimum, and updating the positions of three wolfs, namely alpha, beta and delta.
Step three: according to the optimal sequence obtained by the gray wolf algorithm, marking each sampling point on a grid map one by one;
step four: calculating an optimal raster path between every two sampling points marked in the raster map by using a D-Lite algorithm to obtain an optimal path from a starting point to a final sampling point;
as shown in fig. 3, the specific process is as follows:
step 401: initializing path cost estimation values rhs(s) value, g(s) value and k of each grid pointmThe value is obtained.
Setting relevant parameters of a D-Lite algorithm according to the acquired lake surface two-dimensional coordinate graph and the optimal sequence calculated by the gray wolf optimization algorithm, wherein each grid point is respectively used as a node; s is a set of path nodes in the topographical map, to which S belongs.
Step 402: from the current node s of the grid mapgoalAnd starting to expand the neighboring grids from the current path point to 8 directions around, and selecting the minimum point of the k(s) value as the next expansion node by comparing the evaluation functions k(s) of all the expanded neighboring nodes.
Current node sgoalThe initial value is a first sampling point; due to the movement of large and medium-sized organisms in the water surface, the change of the height of the water surface causes the appearance of rocks, dynamic obstacles are arranged in the map, and the probability of the obstacles appearing on the water surface and the probability of the obstacles disappearing are both set to be 3%, namely the probability that each blank grid becomes an obstacle grid and the probability that the obstacle grid becomes a blank grid is set to be 3%. In the detection process, the cruise ship adopts a laser detector, and the detection distance is 5 grids by 5 grids.
Wherein the evaluation function k(s) comprises two values: k is a radical of1(s) and k2(s), preferentially comparing k1(s) size, when k1When the(s) values are equal, k is compared2(s) size, formula as follows:
k1(s)＝min(g(s),rhs(s))+h(sstart,S)+km
k2(s)＝min(g(s),rhs(s))
in the formula, h(s)startS) is a heuristic function and represents the path cost from the starting point to the current node; k is a radical ofmThe m variable updated when the environment changes;
step 403: taking the next expansion node as the current node, and continuing to expand the adjacent nodes until reaching the node sstartTo obtain a strip fromNode sgoalTo node sstartA path between;
node sstartThe initial value is the initial point of the grid map;
step 404: for the current node sstartCalculating rhs(s) values of all subsequent nodes of the node, and selecting the node with the minimum rhs(s) as a new starting point s'startCruise ship moves to this point s'start。
Step 405: scanning a map, updating node information, judging whether the surrounding environment changes or not, and entering step 406 if the surrounding environment does not change; otherwise, go to step 407;
step 406, moving a cruise ship to a new node s'startAs the current node, the process returns to step 402 again to obtain the slave node sgoalTo the path between the nodes and continue to move the position of the cruise ship until the cruise ship reaches the node sgoalStep 408 is entered;
step 407, updating the variable k in the estimation function k(s)mValue and node s'startReturning to step 402;
the expression is as follows:
km＝km-1+h(slast,s'start)
slast＝s'start
in the formula, km-1Is the m-1 th variable updated when the environment changes, wherein k0＝0；slastIndicating a node, h(s), in front of the cruise shiplast,s'start) Representing one position point ahead of the cruise ship to the current node s'startThe path cost of (2).
If the surrounding environment is not changed in the moving process of the navigation ship, the moving path of the navigation ship and the slave node sgoalTo node sstartExpanded path anastomosis therebetween;
when the surrounding environment changes, the value of the valuation function changes, the extension node of each node also changes, and thus the node sgoalTo node sstartThe path extending between them may also change,the sampling path of the cruise vessel may also change dynamically.
Step 408: judging node sgoalAnd if so, terminating the algorithm to obtain an optimal path traversing a plurality of sampling points. Otherwise, the node s is connectedgoalIs set as a new starting point s'startSetting the next sampling point as the target point sgoalReturning to step 402;
step five: the autonomous cruise ship starts cruise from the starting point corresponding to the grid map, sequentially traverses each sampling point obtained by the wolf optimization algorithm along the optimal grid path, and finally returns to the starting point to finish cruise.
Claims (1)
1. The unmanned cruise ship multi-target path planning method under the dynamic barrier is characterized by comprising the following specific steps:
the method comprises the following steps: the unmanned aerial vehicle collects images of the lake surface environment, and the images are segmented by adopting a grid method to obtain a grid map and set a starting point and d sampling points;
the method specifically comprises the following steps: firstly, dividing a lake surface image into grids with equal size to form a grid map; setting an obstacle in the environment as a black grid, filling irregular obstacles, setting a feasible region as a white grid, and setting a starting point and d sampling points in a grid map by using two-dimensional coordinates;
d is a positive integer;
step two: carrying out sequential optimization on the d sampling points by adopting an improved wolf optimization algorithm to obtain an optimal sequence and storing the optimal sequence;
the steps of obtaining the optimal sequence by adopting the improved wolf optimization algorithm are as follows:
step 201: initializing parameters of a gray wolf optimization algorithm;
the parameters comprise:
the number n of the grey wolfs population and the maximum iteration time t, and the number of sampling points is defined as the dimension d of a search space, the initial position of the grey wolf population and a random path of each grey wolf;
step 202: searching a prey in the space of d sampling points by each wolf, and constructing a space domain matrix P of the position relation between the search space and the wolf;
first, define the location X of the ith wolfiIs a group of positive integer sequences different from each other, and the formula is as follows:
then
Represents the value of the ith grey wolf in the grey wolf population on the d sampling point;
then, the positions of n wolfs constitute a spatial domain matrix P, the formula is as follows:
in the formula (I), the compound is shown in the specification,
the position of the ith wolf at the mth sampling point is shown, and each row in the spatial domain matrix P represents the path of the wolf;
step 203: constructing a distance matrix W to represent the distance between each sampling point and other sampling points;
the formula is as follows:
s(n,m)the distance between the nth sampling point and the mth sampling point is obtained;
step 204: constructing a fitness function f by using the space domain matrix P and the distance matrix W;
the expression is as follows:
Pirefers to the ith row of wolf-only path in the spatial domain matrix P; sigma si(n,m)In the ith route of the wolf, the distances between every two sampling points are added and summed;
step 205: solving a fitness function by using a gray wolf optimization algorithm to obtain an optimal traversal sequence;
the specific process is as follows:
firstly, the searching process:
in the t-generation search process, the position of the gray wolf is X (t), and the position of the prey is XP(t), the distance D of the wolf from the prey is expressed as:
t denotes the current number of iterations, r2Is a random number from 0 to 1; c is a radical of r2A coefficient of variation;
then the bounding process:
in the process that the gray wolf colony surrounds the prey, a relationship model of the gray wolf and the prey is constructed, and the formula is as follows:
wherein A and D are the surrounding step length, a is the improved convergence factor, r1Is a random number from 0 to 1, tmaxDenotes the maximum number of iterations, μ1And mu2Is a constant, λ1And λ2B is a nonlinear continuous function Beta function for adjusting the coefficient;
the improved curve of the convergence factor a is nonlinearly reduced along with the increase of the iteration times, the convergence factor a is nonlinearly and rapidly reduced when the iteration is just started, and the reduction curve of the convergence factor a is approximate to linearity along with the increase of the iteration times;
next, the location update procedure:
according to the relation model of gray wolf and prey, the final positions X of three wolfs of alpha, beta and delta are respectively obtained1，X2And X3Under the guidance of alpha, beta and delta, the whole wolf group gradually approaches to the prey, and the target position is continuously determined through the position updating of the alpha, beta and delta wolfs, so that the final attack position of the wolf on the prey is realized;
in the formula, Xα，Xβ，XδRepresents the updated positions of three wolfs of alpha, beta and delta, A1，A2，A3Three random numbers are represented, and X represents the final attack position of the wolf on the prey; a. the1·Dα，A2·Dβ，A3·DδRepresents the surrounding step sizes of three wolfs of alpha, beta and delta;
finally, judging whether the maximum iteration times is reached, if so, outputting the optimal sequence in the iteration process; otherwise, sequentially storing three optimal solutions obtained when the fitness function is minimum, and updating the positions of three wolfs, namely alpha, beta and delta;
step three: according to the optimal sequence obtained by the gray wolf algorithm, marking each sampling point on a grid map one by one;
step four: calculating an optimal raster path between every two sampling points marked in the raster map by using a D-Lite algorithm to obtain an optimal path from a starting point to a final sampling point;
the specific process is as follows:
step 401: initializing path cost estimation values rhs(s) value, g(s) value and k of each grid pointmA value;
setting relevant parameters of a D-Lite algorithm according to the acquired lake surface two-dimensional coordinate graph and the optimal sequence calculated by the gray wolf optimization algorithm, wherein each grid point is respectively used as a node; s is a set of path nodes in the topographic map, and S belongs to the set S;
step 402: from the current node s of the grid mapgoalStarting to expand neighboring grids from the current path point to 8 directions around, and selecting a minimum point of a value k(s) as a next expansion node by comparing evaluation functions k(s) of all expansion neighboring nodes;
current node sgoalThe initial value is a first sampling point; the method comprises the steps that due to the movement of large and medium-sized organisms in a water surface, rocks appear due to the change of the height of the water surface, dynamic obstacles are arranged in a map, and the probability that the obstacles appear on the water surface and the probability that the obstacles disappear are both 3%, namely the probability that each blank grid becomes an obstacle grid and the probability that the obstacle grid becomes a blank grid is 3%; in the detection process, the cruise ship adopts a laser detector, and the detection distance is 5 grids by 5 grids;
wherein the evaluation function k(s) comprises two values: k is a radical of1(s) and k2(s), preferentially comparing k1(s) size, when k1When the(s) values are equal, k is compared2(s) size, formula as follows:
k1(s)＝min(g(s),rhs(s))+h(sstart,S)+km
k2(s)＝min(g(s),rhs(s))
in the formula, h(s)startS) is a heuristic function and represents the path cost from the starting point to the current node; k is a radical ofmThe m variable updated when the environment changes;
step 403: taking the next expansion node as the current node, and continuing to expand the adjacent nodes until reaching the node sstartTo obtain a slave node sgoalTo node sstartA path between;
node sstartThe initial value is the initial point of the grid map;
step 404: for the current node sstartCalculating rhs(s) values of all subsequent nodes of the node, and selecting the node with the minimum rhs(s) as a new starting point s'startCruise ship moves to this point s'start；
Step 405: scanning a map, updating node information, judging whether the surrounding environment changes or not, and entering step 406 if the surrounding environment does not change; otherwise, go to step 407;
step 406, moving a cruise ship to a new node s'startAs the current node, the process returns to step 402 again to obtain the slave node sgoalTo the path between the nodes and continue to move the position of the cruise ship until the cruise ship reaches the node sgoalStep 408 is entered;
step 407, updating the variable k in the estimation function k(s)mValue and node s'startReturning to step 402;
the expression is as follows:
km＝km-1+h(slast,s'start)
slast＝s'start
in the formula, km-1Is the m-1 th variable updated when the environment changes, wherein k0＝0；slastIndicating a node, h(s), in front of the cruise shiplast,s'start) Representing one position point ahead of the cruise ship to the current node s'startThe path cost of (2);
if the surrounding environment is not changed in the moving process of the navigation ship, the moving path of the navigation ship and the slave node sgoalTo node sstartExpanded path anastomosis therebetween;
when the surrounding environment changes, the value of the valuation function changes, the extension node of each node also changes, and thus the node sgoalTo node sstartThe extended path between the two paths can be changed, and the sampling path of the cruise ship can be dynamically changed;
step 408: judging node sgoalWhether the sampling point is the final sampling point or not is judged, if so, the algorithm is terminated, and an optimal path traversing a plurality of sampling points is obtained; otherwise, the node s is connectedgoalIs set as a new starting point s'startSetting the next sampling point as the target point sgoalReturning to step 402;
step five: and the autonomous cruise ship starts cruising from the starting point corresponding to the grid map, sequentially passes through each sampling point along the optimal grid path and finally returns to the starting point to finish cruising.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010340555.6A CN111457927B (en) | 2020-04-26 | 2020-04-26 | Unmanned cruise ship multi-target path planning method under dynamic barrier |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010340555.6A CN111457927B (en) | 2020-04-26 | 2020-04-26 | Unmanned cruise ship multi-target path planning method under dynamic barrier |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111457927A CN111457927A (en) | 2020-07-28 |
| CN111457927B true CN111457927B (en) | 2021-03-05 |
Family
ID=71683856
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010340555.6A Active CN111457927B (en) | 2020-04-26 | 2020-04-26 | Unmanned cruise ship multi-target path planning method under dynamic barrier |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111457927B (en) |
Families Citing this family (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112013841B (en) * | 2020-08-26 | 2021-08-06 | 南京工业大学 | Semantic SLAM service robot navigation method in indoor dynamic environment |
| CN112113571B (en) * | 2020-09-18 | 2021-08-17 | 武汉理工大学 | Planning method for coverage paths of multiple unmanned measurement boats |
| CN112113573B (en) * | 2020-09-18 | 2021-08-10 | 武汉理工大学 | Planning method for coverage path of single unmanned measurement boat |
| CN112130587A (en) * | 2020-09-30 | 2020-12-25 | 北京理工大学 | Multi-unmanned aerial vehicle cooperative tracking method for maneuvering target |
| CN112256029A (en) * | 2020-10-16 | 2021-01-22 | 中冶赛迪上海工程技术有限公司 | Control method and system for unmanned steel grabbing machine |
| CN112288813B (en) * | 2020-11-03 | 2022-06-21 | 浙江大学 | Pose estimation method based on multi-view vision measurement and laser point cloud map matching |
| CN112747736B (en) * | 2020-12-22 | 2022-07-08 | 西北工业大学 | Indoor unmanned aerial vehicle path planning method based on vision |
| CN112799405B (en) * | 2021-01-05 | 2022-06-28 | 北京工商大学 | Unmanned ship path planning method based on dynamic barrier environment |
| CN113203420B (en) * | 2021-05-06 | 2022-11-18 | 浙江大学 | Industrial robot dynamic path planning method based on variable density search space |
| CN113580129B (en) * | 2021-07-19 | 2024-01-16 | 中山大学 | Multi-target cooperative capturing method, device and medium based on robot |
| CN113759915B (en) * | 2021-09-08 | 2023-09-15 | 广州杰赛科技股份有限公司 | AGV trolley path planning method, device, equipment and storage medium |
| CN114219992B (en) * | 2021-12-14 | 2022-06-03 | 杭州古伽船舶科技有限公司 | Unmanned ship obstacle avoidance system based on image recognition technology |
| CN115248591B (en) * | 2021-12-28 | 2023-06-09 | 齐齐哈尔大学 | UUV path planning method based on mixed initialization wolf particle swarm algorithm |
| CN114115301B (en) * | 2022-01-26 | 2022-04-22 | 河北工业大学 | Mobile robot improved A-algorithm based on wolf colony algorithm and artificial potential field |
| CN114462627A (en) * | 2022-03-16 | 2022-05-10 | 兰州理工大学 | Method for diagnosing abnormity of top-blown smelting system based on Hui wolf algorithm and support vector machine |
| CN114357106A (en) * | 2022-03-21 | 2022-04-15 | 广东省科学院广州地理研究所 | Indoor path planning method and device |
| CN114973443B (en) * | 2022-05-19 | 2024-04-12 | 杭州中威电子股份有限公司 | Inspection robot-based complex gas environment inspection system and method thereof |
| CN115333990B (en) * | 2022-10-14 | 2023-02-03 | 中国人民解放军海军工程大学 | Efficient transmission routing method based on suburb and gray wolf hybrid optimization |
| CN115437386B (en) * | 2022-11-03 | 2023-02-24 | 中国人民解放军陆军装甲兵学院 | Unmanned vehicle route planning method based on air-ground information fusion |
| CN115826587B (en) * | 2023-02-14 | 2023-06-06 | 河北工业大学 | Mobile robot path planning method and device based on road network ant colony algorithm |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6851241B2 (en) * | 2001-01-12 | 2005-02-08 | Valinge Aluminium Ab | Floorboards and methods for production and installation thereof |
| JP2012205501A (en) * | 2011-03-27 | 2012-10-22 | Boeing Co:The | Sequential shunt regulator with analog fill control |
| CN107153889A (en) * | 2017-05-03 | 2017-09-12 | 北京工商大学 | Water quality sampling cruise ship path planning optimal method |
| CN110244720A (en) * | 2019-06-04 | 2019-09-17 | 浙江海洋大学 | Paths planning method and system for marine unmanned boat |
| CN110276437A (en) * | 2019-06-21 | 2019-09-24 | 广西大学 | Mixing based on Fuch mapping improves grey wolf optimization algorithm |
| CN111061273A (en) * | 2019-12-26 | 2020-04-24 | 北京航天控制仪器研究所 | Autonomous obstacle avoidance fusion method and system for unmanned ship |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080285640A1 (en) * | 2007-05-15 | 2008-11-20 | Crestcom, Inc. | RF Transmitter With Nonlinear Predistortion and Method Therefor |
| EP2804334A1 (en) * | 2013-05-13 | 2014-11-19 | Xieon Networks S.à.r.l. | Method, device and communication system for reducing optical transmission impairments |
| CN108183500B (en) * | 2017-11-24 | 2020-07-10 | 国网甘肃省电力公司电力科学研究院 | Multi-energy complementary rural micro-energy network capacity optimization configuration method and device |
| CN109242139A (en) * | 2018-07-23 | 2019-01-18 | 华北电力大学 | A kind of electric power day peak load prediction technique |
| CN111024433A (en) * | 2019-12-30 | 2020-04-17 | 辽宁大学 | Industrial equipment health state detection method for optimizing support vector machine by improving wolf algorithm |
-
2020
- 2020-04-26 CN CN202010340555.6A patent/CN111457927B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6851241B2 (en) * | 2001-01-12 | 2005-02-08 | Valinge Aluminium Ab | Floorboards and methods for production and installation thereof |
| JP2012205501A (en) * | 2011-03-27 | 2012-10-22 | Boeing Co:The | Sequential shunt regulator with analog fill control |
| CN107153889A (en) * | 2017-05-03 | 2017-09-12 | 北京工商大学 | Water quality sampling cruise ship path planning optimal method |
| CN110244720A (en) * | 2019-06-04 | 2019-09-17 | 浙江海洋大学 | Paths planning method and system for marine unmanned boat |
| CN110276437A (en) * | 2019-06-21 | 2019-09-24 | 广西大学 | Mixing based on Fuch mapping improves grey wolf optimization algorithm |
| CN111061273A (en) * | 2019-12-26 | 2020-04-24 | 北京航天控制仪器研究所 | Autonomous obstacle avoidance fusion method and system for unmanned ship |
Non-Patent Citations (12)
| Title |
|---|
| "A Hybrid Path Planning Method for an Unmanned Cruise Ship in Water Quality Sampling";Yu, Jiabin等;《IEEE ACCESS》;20191231;第7卷;87127-87140 * |
| "A novel hybrid grey wolf optimizer algorithm for unmanned aerial vehicle (UAV) path planning";Qu, Chengzhi等;《KNOWLEDGE-BASED SYSTEMS》;20200422;第194卷;全文 * |
| "Grey Wolf Optimization for Global Path Planning of Autonomous Underwater Vehicle";Panda, Madhusmita等;《PROCEEDINGS OF THE THIRD INTERNATIONAL CONFERENCE ON ADVANCED INFORMATICS FOR COMPUTING RESEARCH (ICAICR "19)》;20191231;全文 * |
| "Three dimensional path planning using Grey wolf optimizer for UAVs";Dewangan, Ram Kishan等;《APPLIED INTELLIGENCE》;20190630;第49卷(第6期);2201-2217 * |
| "基于D*Lite算法的移动机器人路径规划研究";徐开放;《中国优秀硕士学位论文全文数据库信息科技辑》;20180215(第02期);3.2.4D*Lite搜索算法 * |
| "基于改进A*算法的无人船完全遍历路径规划";吕霞付等;《水下无人系统学报》;20191231;第27卷(第6期);1基于栅格法的USV环境建模，2基于栅格地图的完全遍历算法 * |
| "基于改进灰狼优化算法的移动机器人路径规划";刘宁宁等;《电测与仪表》;20200110;第57卷(第1期);76-83,98 * |
| "基于改进灰狼优化算法的类ＴＳＰ问题研究—以旅游为例";许如琪等;《地理与地理信息科学》;20180331;第３４卷(第２期);０引言，１算法原理 * |
| "巡航船污染水质采集路径规划仿真研究";宗陈等;《计算机仿真》;20180930;第35卷(第9期);338-342,390 * |
| "改进多目标蚁群算法在动态路径优化中的应用";吴耕锐等;《计算机应用与软件》;20190531;第36卷(第5期);249-254,288 * |
| "旅行商问题(TSP)的现代优化算法研究";蔡晨晓等;《舰船电子工程》;20081231(第174期);114-117 * |
| "贪婪随机自适应灰狼优化算法求解TSP问题";高珊等;《现代电子技术》;20190715;第42卷(第14期);46-50,54 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111457927A (en) | 2020-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111457927B (en) | Unmanned cruise ship multi-target path planning method under dynamic barrier | |
| CN110632931A (en) | Mobile robot collision avoidance planning method based on deep reinforcement learning in dynamic environment | |
| CN108388250B (en) | Water surface unmanned ship path planning method based on self-adaptive cuckoo search algorithm | |
| CN112987799B (en) | Unmanned aerial vehicle path planning method based on improved RRT algorithm | |
| CN107229287A (en) | A kind of unmanned plane global path planning method based on Genetic Ant algorithm | |
| CN113486764B (en) | Pothole detection method based on improved YOLOv3 | |
| CN112911705A (en) | Bayesian iteration improved particle swarm optimization algorithm-based indoor positioning method | |
| CN112799405B (en) | Unmanned ship path planning method based on dynamic barrier environment | |
| CN112033413A (en) | Improved A-algorithm combined with environmental information | |
| CN111721296B (en) | Data driving path planning method for underwater unmanned vehicle | |
| CN113848919A (en) | Ant colony algorithm-based indoor AGV path planning method | |
| CN114237235B (en) | Mobile robot obstacle avoidance method based on deep reinforcement learning | |
| CN110906935A (en) | Unmanned ship path planning method | |
| CN112000126B (en) | Whale algorithm-based multi-unmanned-aerial-vehicle collaborative searching multi-dynamic-target method | |
| CN114167865A (en) | Robot path planning method based on confrontation generation network and ant colony algorithm | |
| CN114659530A (en) | Grid model map construction method for intelligent robot path planning | |
| CN113064422B (en) | Autonomous underwater vehicle path planning method based on double neural network reinforcement learning | |
| CN115248591A (en) | UUV path planning method based on hybrid initialization Hui wolf particle swarm algorithm | |
| Qiao et al. | Sample-based frontier detection for autonomous robot exploration | |
| CN116700284A (en) | AGV path planning method based on improved self-adaptive weight A-based algorithm | |
| CN115619953A (en) | Rugged terrain-oriented mobile robot terrain mapping method and system | |
| Xue et al. | Real-time 3D grid map building for autonomous driving in dynamic environment | |
| CN115129064A (en) | Path planning method based on fusion of improved firefly algorithm and dynamic window method | |
| CN115496998A (en) | Remote sensing image wharf target detection method | |
| CN114153216A (en) | Lunar surface path planning system and method based on deep reinforcement learning and block planning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |